Lessons learned? [Regulatives / Guidelines]

posted by Helmut Homepage – Vienna, Austria, 2011-08-24 15:51 (4600 d 11:37 ago) – Posting: # 7304
Views: 6,485

Dear control freaks!

❝ Second I come more and more to the conclusion that a decomposition of the subject effect into the terms sequence and subject nested within sequence is not really possible with those crippled datasets […]. The robust model without any peculiarities should therefore include period and subject effects only.

❝ Do I err here?


If you want to get a second amateur’s opinion: exactly. That’s why I called the datasets (after removal of T!) ‘crippled’. All information about sequences (RTRT ^ TRTR ¬ RR!) is lost. Martin (not an amateur) once told me that you can do only statistics if you understand the data generating process. In all replicate datasets I have seen until now, EMA’s method gives a smaller CVWR than the full model (= FDA’s or EMA’s ‘Method C’). Politics = narrower AR? If we throw away sequences, it doesn’t make any sense to include them in the model (not to speak about nested subjects either).

BTW, more than 20 years after Freeman’s paper1, Senn’s book2, and the empiric study by D’Angelo et al.3 EMA has realised (see IR GL) that sequence effects are of no importance in a properly designed cross-over study. But: Would anybody evaluate a 2×2 cross-over by Y=treatment+period+subject only - though the CI stays the same?

Agree with your third point.


[image]Quotes on a sad occasion:
  1. Freeman PR. The Performance of the Two-Stage Analysis of Two-Treatment, Two-Period Crossover Trials. Stat Med. 1989;8:1421–32.
  2. S Senn S. Cross-over Trials in Clinical Research. Chichester: John Wiley & Sons; 2nd ed. 2002.
  3. D’angelo G, Potvin D, Turgeon J. Carry-Over Effects in Bioequivalence Studies. J Biopharm Stat. 2001;11(1&2):35–43.

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

Complete thread:

UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
75 visitors (0 registered, 75 guests [including 8 identified bots]).
Forum time: 02:28 CET (Europe/Vienna)

With four parameters I can fit an elephant,
and with five I can make him wiggle his trunk.    John von Neumann

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5